Detection method based on detector wireless signal and detector device

Abstract

The invention relates to the technical field of signal detection, and discloses a detection method based on a detector wireless signal and a detector device.The method comprises the steps that a wireless signal receiving power value and a signal source distance value are obtained through a detector; performing channel attenuation compensation on the wireless signal receiving power value based on the signal source distance value to obtain a wireless signal transmitting power value; performing signal envelope detection and transmission feature recognition on wireless signals in the environment to obtain a camera transmission feature set; carrying out probability calculation on a camera signal state and a non-camera signal state to obtain a signal state transition probability value; according to the hidden camera detection method, the time consistency and the stability of the detection result are ensured, compared with a traditional method, the false alarm rate and the missing report rate are remarkably reduced, and the accuracy and the reliability of hidden camera detection are improved.

Description

Technical Field

[0001] This invention relates to the field of signal detection technology, and in particular to a detection method and detector device based on wireless signals from a detector. Background Technology

[0002] Traditional methods for detecting hidden cameras primarily rely on optical detection, infrared detection, or simple radio frequency signal detection. However, in complex electromagnetic environments, signal interference from wireless devices such as Wi-Fi routers, Bluetooth devices, and mobile phones severely impacts the accuracy of detectors. Existing wireless signal detection technologies typically employ fixed power thresholds or simple signal strength judgment methods, ignoring the attenuation characteristics of signals during propagation and the influence of environmental factors. They fail to recognize the unique burst transmission patterns of camera signals, resulting in the detection system's inability to distinguish camera signals from signals from other wireless devices. Summary of the Invention

[0003] This invention provides a detection method and detector device based on the wireless signal of the detector. This invention ensures the temporal consistency and stability of the detection results, and significantly reduces the false alarm rate and false negative rate compared with traditional methods, thereby improving the accuracy and reliability of hidden camera detection.

[0004] In a first aspect, the present invention provides a detection method based on detector wireless signals, the detection method based on detector wireless signals comprising: The detector acquires the wireless signal reception power value and the signal source distance value; Based on the signal source distance value, channel attenuation compensation is performed on the wireless signal received power value to obtain the wireless signal transmitted power value; Signal envelope detection and transmission feature identification are performed on wireless signals in the environment to obtain a set of camera transmission features; Based on the set of camera transmission features, perform probability calculations on camera signal states and non-camera signal states to obtain signal state transition probability values; The hidden camera detection result is generated by weighted fusion decision based on the signal state transition probability value and the wireless signal transmission power value.

[0005] In conjunction with the first aspect, in a first implementation of the first aspect of the present invention, the step of acquiring the wireless signal received power value and the signal source distance value through the detector includes: The system synchronously receives wireless signals in the 2.4GHz and 5GHz bands in the environment to obtain multi-band camera wireless signal data and corresponding wireless signal received power values, while simultaneously measuring the received signal strength indicator values ​​for each band. The detector acquires ambient temperature, relative humidity, and spatial obstacle density values ​​based on its built-in environmental parameter sensor array, and calculates the current environmental loss factor value. The signal phase difference analysis and signal propagation time difference analysis are performed on the wireless signal data of the multi-band camera, and the propagation delay is corrected by combining the current environmental loss factor value to obtain the signal arrival angle value and the signal propagation time difference value. Three-dimensional spatial geometric positioning is performed based on the camera signal arrival angle value and the signal propagation time difference value to obtain the signal source distance value in the environment.

[0006] In conjunction with the first aspect, in a second implementation of the first aspect of the present invention, the step of performing channel attenuation compensation on the wireless signal received power value based on the signal source distance value to obtain the wireless signal transmitted power value includes: The signal propagation path loss value is obtained by performing a logarithmic operation on the free space path loss based on the signal source distance value. Based on the current environmental loss factor value, the environmental-related attenuation loss value is obtained by weighted calculation using the environmental temperature correction coefficient, relative humidity correction coefficient, and spatial obstacle correction coefficient. The total channel attenuation loss is obtained by summing the signal propagation path loss value and the environment-related attenuation loss value. The total channel attenuation loss value is added to the wireless signal received power value to obtain the wireless signal transmitted power value.

[0007] In conjunction with the first aspect, in a third implementation of the first aspect of the present invention, the step of performing signal envelope detection and transmission feature identification on wireless signals in the environment to obtain a set of camera transmission features includes: The signal envelope of the wireless signal in the environment is extracted to obtain the wireless signal envelope waveform data; Based on the wireless signal envelope waveform data, the time distribution of the active and silent segments of the signal is identified to obtain the time series data of the active segments. The time-domain parameters of the camera burst transmission mode are statistically analyzed on the time series data of the signal activity segment. The duration of the continuous signal activity segment is calculated as the transmission period parameter, the interval time between adjacent activity segments is calculated as the silence interval parameter, and the data packet length parameter is calculated by modulation characteristics and bit rate analysis to obtain the combination of camera time-domain characteristic parameters. A set of camera transmission features is constructed based on the combination of camera temporal feature parameters and power variance parameters.

[0008] In conjunction with the first aspect, in a fourth implementation of the first aspect of the present invention, the step of identifying the time distribution of active and silent segments of the signal based on the wireless signal envelope waveform data to obtain time series data of the active signal segments includes: Calculate the mean and standard deviation of the wireless signal envelope waveform data respectively; The signal activity segment detection threshold is set by adding the mean value to the product of the adjustment coefficient and the standard deviation value; Based on the signal activity segment detection threshold, the amplitude determination calculation is performed on the wireless signal envelope waveform data point by point. When the envelope amplitude exceeds the detection threshold, it is identified as a signal activity state, and when it is below the detection threshold, it is identified as a signal silence state, thus obtaining the signal activity segment time series data.

[0009] In conjunction with the first aspect, in a fifth implementation of the first aspect of the present invention, the step of performing probability calculations of camera signal states and non-camera signal states based on the camera transmission feature set to obtain signal state transition probability values ​​includes: The probability density function of the camera signal state and the probability of the first camera signal state are calculated by using a binary state transition probability model on the camera transmission feature set to obtain the probability of the first camera signal state and the probability of the first non-camera signal state. Based on the state probability of the first camera signal and the state probability of the first non-camera signal, a Bayesian posterior probability update operation is performed by combining the state probability at the current moment and the Markov chain transformation matrix to obtain the state probability of the second camera signal and the state probability of the second non-camera signal. Signal state transition probability values ​​are generated based on the second camera signal state probability and the second non-camera signal state probability.

[0010] In conjunction with the first aspect, in the sixth implementation of the first aspect of the present invention, the step of calculating the probability density function of the camera signal state and the non-camera signal state of the camera transmission feature set using a binary state transition probability model to obtain the first camera signal state probability and the first non-camera signal state probability includes: The camera transmission feature set is input into the Gaussian mixture model corresponding to the camera signal state and the non-camera signal state in the binary state transition probability model to calculate the probability density function, and the result of the dual-state Gaussian mixture probability density calculation is obtained. Based on the calculated results of the dual-state Gaussian mixture probability density, the probability density values ​​under the camera signal state and the probability density values ​​under the non-camera signal state are calculated respectively to obtain the probability of the first camera signal state and the probability of the first non-camera signal state.

[0011] In conjunction with the first aspect, in the seventh implementation of the first aspect of the present invention, the step of performing a Bayesian posterior probability update operation based on the first camera signal state probability and the first non-camera signal state probability, combined with the current state probability and the Markov chain transformation matrix, to obtain the second camera signal state probability and the second non-camera signal state probability, includes: Obtain the prior probability of the camera signal state and the prior probability of the non-camera signal state at the current moment from the historical state records; Based on the prior probabilities of the camera signal state and the prior probabilities of the non-camera signal state, the transition probabilities from camera signal state to camera signal state, from camera signal state to non-camera signal state, from non-camera signal state to camera signal state, and from non-camera signal state to non-camera signal state are extracted from the Markov chain transition matrix to obtain a quaternary state transition probability matrix. The probability of the first camera signal state and the probability of the first non-camera signal state are used as likelihood probabilities. Bayesian operations are then performed on the quaternary state transition probability matrix to obtain the probability of the second camera signal state and the probability of the second non-camera signal state.

[0012] In conjunction with the first aspect, in the eighth implementation of the first aspect of the present invention, the step of generating a hidden camera detection result by weighted fusion decision based on the signal state transition probability value and the wireless signal transmission power value includes: The product of the signal state transition probability value and the first weighting coefficient is added to the product of the wireless signal transmission power value and the second weighting coefficient to obtain the comprehensive fusion decision score. The integrated decision score is compared with the camera detection decision threshold, and the probability decision value of the existence of the hidden camera signal is output. The probability judgment value of the presence of the hidden camera signal is verified for continuity within a time window. The frequency of the camera signal presence judgment within the time window is counted. When the frequency exceeds the preset consistency verification threshold, the hidden camera detection result is confirmed.

[0013] In a second aspect, the present invention provides a detector device, the detector device comprising: The acquisition module is used to acquire the wireless signal received power value and the signal source distance value through the detector; The compensation module is used to perform channel attenuation compensation on the wireless signal received power value based on the signal source distance value to obtain the wireless signal transmitted power value; The identification module is used to detect the signal envelope and identify the transmission features of wireless signals in the environment to obtain a set of camera transmission features. The calculation module is used to perform probability calculations of camera signal state and non-camera signal state based on the camera transmission feature set, and obtain signal state transition probability values; The generation module is used to perform weighted fusion decision based on the signal state transition probability value and the wireless signal transmission power value to generate the hidden camera detection result.

[0014] The technical solution provided by this invention establishes a channel attenuation compensation model that includes free space path loss and environmental loss factors, enabling accurate reconstruction of the true transmission power of wireless signals and eliminating the impact of distance and environmental factors on detection accuracy. Employing synchronous reception technology in the 2.4GHz and 5GHz bands, combined with signal phase difference analysis and time difference analysis of a three-antenna array, precise three-dimensional positioning and full-band coverage of the signal source are achieved. By extracting the unique burst transmission patterns of camera signals, including time-domain characteristic parameters such as transmission period, silence interval, and data packet length, camera signals can be effectively distinguished from signals from other wireless devices such as WiFi and Bluetooth. A binary state transition probability model using a Markov chain framework and Bayesian inference algorithm establishes an intelligent probabilistic inference mechanism, achieving comprehensive utilization of multi-dimensional information through weighted fusion decision technology. Combined with continuous verification within a sliding time window, the temporal consistency and stability of the detection results are ensured, significantly reducing the false alarm rate and false negative rate compared to traditional methods, and improving the accuracy and reliability of hidden camera detection. Attached Figure Description

[0015] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0016] Figure 1 This is a schematic diagram of the steps of the detection method based on the detector's wireless signal in an embodiment of the present invention; Figure 2 This is a schematic diagram of the detector device in an embodiment of the present invention. Detailed Implementation

[0017] This invention provides a detection method and detector device based on a detector's wireless signal. The terms "first," "second," "third," "fourth," etc. (if present) in the specification, claims, and accompanying drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" or "having" and any variations thereof are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or device that includes a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or devices.

[0018] For ease of understanding, the specific process of the embodiments of the present invention is described below. Please refer to [link / reference]. Figure 1 One embodiment of the detection method based on detector wireless signals in this invention includes: Step S1: Obtain the wireless signal received power value and signal source distance value through the detector; It is understood that the executing entity of this invention can be a detector device, a terminal, or a server; no specific limitation is made here. This embodiment of the invention will be described using a server as an example.

[0019] Specifically, the detector activates its software-defined radio module, which has multi-band reception capabilities, and simultaneously collects and processes wireless signals in the 2.4 GHz and 5 GHz bands in the environment during reception. By covering these two main frequency bands, the detector simultaneously obtains multi-band wireless signal data from the hidden camera, and combines this with the received power value measured by the receiving antenna array in each band. Simultaneously, it extracts the RSSI (Received Signal Strength Indicator) value, which is directly related to the received power, in real time to establish a preliminary judgment of the channel conditions. The detector relies on its internally configured environmental parameter sensor array to continuously monitor key physical parameters in the current environment, including ambient temperature, relative humidity, and the density of obstacles in the space. These parameters, as key variables affecting signal propagation characteristics, are used to construct a dynamic environmental loss factor. This loss factor not only considers free-space path loss but also incorporates the superimposed effects of temperature and humidity on signal absorption attenuation, as well as the additional attenuation caused by multipath interference from obstacles, making the wireless propagation model in specific scenarios closer to actual propagation conditions. In signal path estimation, a multi-antenna receiving device is used to calculate the phase difference of the collected multi-band wireless signal data. The incident direction of the signal wavefront is then deduced from the phase information received by each antenna. Simultaneously, the time difference of signal arrival is analyzed using timestamp data from each receiving channel to obtain the propagation delay difference. Since signal propagation speed in the actual environment is affected by factors such as temperature, humidity, and obstacle materials, the propagation delay is corrected based on the previously calculated environmental loss factor, thereby eliminating systematic biases in propagation time estimation caused by environmental changes. Based on the camera signal arrival angle and signal propagation time difference, a three-dimensional geometric positioning method is used, combining the spatial deployment parameters of multiple antennas with the corrected measurement values ​​to construct a three-dimensional spatial positioning equation set. Solving this equation set determines the azimuth and elevation angles of the signal source in the detector's reference coordinate system, and the accurate distance from the detector to the signal source is calculated.

[0020] Step S2: Perform channel attenuation compensation on the wireless signal received power value based on the signal source distance value to obtain the wireless signal transmitted power value; Specifically, based on the signal source distance, a path loss estimation mechanism based on a free-space propagation model is constructed, and this mechanism is used to calculate the power attenuation of the signal during propagation. The calculation of this path loss relies on a logarithmic free-space loss model. By substituting the signal source distance into the free-space path loss formula and performing a logarithmic operation, the signal propagation path loss value is obtained. This value reflects the energy attenuation of electromagnetic waves due to spatial diffusion at a corresponding distance under ideal conditions. Simultaneously, to improve the adaptability and accuracy of the compensation calculation to the actual environment, multiple correction dimensions are introduced based on the environmental loss factor values ​​obtained from the current detector, including environmental temperature correction coefficient, relative humidity correction coefficient, and spatial obstacle correction coefficient. By weighting these three correction coefficients with their corresponding environmental parameters and superimposing them with the original environmental loss baseline value, the environmentally correlated attenuation loss value is obtained. This value comprehensively reflects the additional attenuation effect of the signal in the current physical environment caused by non-free-space factors such as weather conditions and obstacle obstruction. The signal propagation path loss value and the environmentally correlated attenuation loss value are summed, and the resulting total value is the total channel attenuation loss value, representing the total loss amplitude of the signal propagating from the transmitter to the detector receiver. Based on this, in order to achieve dynamic restoration of signal transmission power, the currently received wireless signal power value and the total channel attenuation loss value are superimposed with power compensation, and the compensation value is directly added to the received power to obtain the original estimated value of wireless signal transmission power.

[0021] Step S3: Perform signal envelope detection and transmission feature recognition on the wireless signals in the environment to obtain the camera transmission feature set; Specifically, based on multi-band wireless signals, envelope extraction is performed on the original signal using time-domain processing. This process employs Hilbert transform as the core analysis method, constructing an analytical signal from the original real-valued signal and extracting its envelope curve to obtain wireless signal envelope waveform data reflecting the change of signal strength over time. Based on this envelope data, a dynamic threshold based on statistical distribution is set to determine whether the envelope amplitude is higher than the background noise level. This threshold is used to classify and identify the waveform, thereby extracting a series of signal activity segments and silent segments. The occurrence time and duration of these segments are recorded in time-series format, constructing signal activity segment time-series data. Parametric analysis is then performed on this time-series data to determine whether the wireless signal exhibits the burst transmission pattern unique to the camera. Specifically, the duration of each continuous activity segment is used as the transmission period parameter, and the time interval between any two adjacent activity segments is calculated as the silent interval parameter. The total number of data bits contained in each activity segment is calculated by back-calculating the data packet length parameter using modulation scheme determination and known bit rate information. These three statistical parameters—transmission period, silent interval, and data packet length—constitute the main expression of the camera's transmission characteristics in the time domain. Variance statistics are performed on the power fluctuations within the active segment, and the power variance parameter is extracted to quantify the power consistency or trend of the signal within the active segment, thereby helping to determine whether the signal belongs to an intermittent device with unstable power consumption characteristics. The above-mentioned time-domain feature parameter set is fused and combined with the power variance parameter to construct a multi-dimensional camera transmission feature set.

[0022] Step S4: Calculate the probability of camera signal state and non-camera signal state based on the camera transmission feature set to obtain the signal state transition probability value; Specifically, a probability density analysis mechanism related to state recognition is established based on the camera transmission feature set. A pre-trained binary state transition probability model is used to estimate the probability density function of the input multidimensional camera transmission feature set, and the feature distribution probability is calculated for two candidate states: "camera signal state" and "non-camera signal state," yielding the probability of the first camera signal state and the probability of the first non-camera signal state. This process models historical feature samples based on a Gaussian mixture model, extracts the mean vector and covariance matrix for each state category, and combines maximum likelihood estimation or expectation-maximization algorithms to perform probability matching on real-time observed features, thereby completing the conditional probability density evaluation of the input features under each category. Based on the probability of the first camera signal state and the probability of the first non-camera signal state, a Markov state transition mechanism based on time series modeling is introduced, combining the known or predicted state probability values ​​from previous moments with the current observed state to construct a dynamic process of state transitions. Based on the established Markov chain state transition matrix, historical state values ​​and current feature observations are updated using a Bayesian inference formula to perform posterior probability updates. In this process, the first-stage probability value is used as the likelihood term, the historical state value as the prior term, and the transition probabilities in the transition matrix as conditional weighting factors, thus outputting the updated probabilities of the second camera signal state and the second non-camera signal state. This Bayesian update process reflects the instantaneous state information of the current observation data and explicitly preserves the temporal continuity of the state evolution process, thereby enhancing the model's adaptability to signal behavior patterns in dynamically changing environments. Based on the second-stage state probability results, signal state transition probability values ​​are constructed to measure the overall probability trend of the target wireless signal transitioning from a non-camera state to a camera state at the current moment.

[0023] Step S5: Perform a weighted fusion decision based on the signal state transition probability value and the wireless signal transmission power value to generate the hidden camera detection result.

[0024] Specifically, two weighting coefficients are set to measure the relative importance of state probability and power information in the final decision process. The first weighting coefficient assigns priority to the signal state transition probability value for behavior recognition, while the second weighting coefficient measures the reference role of signal power in device identification. The signal state transition probability value is multiplied by the first weighting coefficient, and the wireless signal transmission power value is multiplied by the second weighting coefficient. The two are then summed to generate a comprehensive fusion decision score. This score numerically reflects the overall trend of the target signal being identified as a hidden camera signal at the current moment. The comprehensive fusion decision score is compared with a preset camera detection decision threshold. If the fusion score exceeds the threshold, a corresponding probability decision value is output, indicating that at that moment, the wireless signal has a high probability of originating from a hidden camera transmitting device, thus marking it as a potential surveillance behavior. To avoid misjudgments caused by noise, transient interference, or environmental fluctuations in a single decision, a time window mechanism is introduced to continuously verify the decision results. In this process, a fixed-length time window is set, and each decision result within the window is statistically accumulated, recording the frequency of "camera signal presence" judgments. This frequency is then compared with a preset consistency verification threshold. When the frequency exceeds the threshold, the determination of the hidden camera signal's presence is confirmed to have temporal continuity and stability, thus outputting the hidden camera detection result. This detection result includes the detection conclusion, along with the signal source's spatial location, the judgment confidence level, and the detection timestamp.

[0025] In one specific embodiment, the process of performing step S1 may specifically include the following steps: The system synchronously receives wireless signals in the 2.4GHz and 5GHz bands in the environment to obtain multi-band camera wireless signal data and corresponding wireless signal received power values, while simultaneously measuring the received signal strength indicator values ​​for each band. The detector acquires ambient temperature, relative humidity, and spatial obstacle density values ​​based on its built-in environmental parameter sensor array, and calculates the current environmental loss factor value. The signal phase difference analysis and signal propagation time difference analysis are performed on the wireless signal data of multi-band cameras, and the propagation delay is corrected by combining the current environmental loss factor value to obtain the signal arrival angle value and the signal propagation time difference value. Three-dimensional spatial geometric positioning is performed based on the camera signal arrival angle and signal propagation time difference to obtain the distance value of the signal source in the environment.

[0026] Specifically, the detector system establishes a parallel acquisition channel through its built-in multi-band receiving module. This module adopts a software-defined radio architecture to achieve synchronous coverage of 2.4GHz and 5GHz frequency bands at the physical level, ensuring the integrity of the frequency domain response even under dynamic channel changes or multi-source signal interference. In this architecture, the system sets up two independent RF front-end links, each configured with a tunable local oscillator, bandpass filter, and low-noise amplifier to achieve RF selection and signal amplification for the corresponding frequency band. Then, digital sampling is performed through a unified intermediate frequency processing platform. The sampling frequency is controlled at the 40MHz level, and the sampling bit depth is 14 bits, thus preserving the detailed changes in signal in both time and amplitude dimensions and ensuring the synchronization between multi-band signals. After signal reception, the system extracts the corresponding received power values ​​for the wireless signals in the 2.4GHz and 5GHz bands respectively, and obtains the average power index using window integration and energy estimation. At the same time, it combines the RSSI (Received Signal Strength Indicator) values ​​of each frequency band for strength calibration to construct a multi-band wireless signal power database. The system utilizes its environmental perception unit, an integrated array of environmental parameter sensors within the detector, to perform real-time sensing and quantitative measurement of the current physical space's environmental conditions. The sensor array includes a high-precision temperature sensor, a capacitive humidity sensor, and a millimeter-wave or ultrasonic obstacle detection module, which collects ambient temperature, relative humidity, and spatial obstacle density parameters, respectively. The system reads the sensor output data and performs unit conversion and filtering averaging to ensure parameter stability and measurement reliability. Based on this, a weighted environmental loss model is introduced, including ambient temperature correction factors, humidity correction factors, and obstacle occlusion factors. These correction terms are multiplied by the deviation values ​​of their corresponding environmental variables in the model and combined with the basic path loss model to generate the environmental loss factor value for the current scenario. This factor reflects the additional attenuation effect under non-ideal propagation conditions. The system performs phase difference analysis and propagation time difference analysis based on the collected multi-band wireless signal data, achieving joint analysis of spatial direction and distance through the signal arrival characteristics between multiple receiving antennas. In the phase difference analysis section, the system deploys at least three equally spaced directional receiving antennas, with a spacing of half a wavelength (λ / 2). The same signal received by different antennas is compared in the time domain, and the phase value of each channel signal at the dominant frequency is extracted using a Fast Fourier Transform (FFT), and the phase difference is calculated. Ideally, the phase difference can be converted into an estimate of the angle of arrival (AOA), calculated by combining the signal wavelength and antenna spacing, using an arcsine function to obtain the signal's incident angle. Simultaneously, the propagation time difference analysis module estimates the time delay by measuring the time difference between the same signal received by different antennas and employing the cross-correlation peak method.To improve measurement accuracy, the system combines phase and time parameter models and uses the current environmental loss factor as a correction parameter to dynamically adjust the propagation speed. This corrects for nonlinear interference caused by environmental factors affecting the propagation path speed and phase rotation, obtaining the converged arrival angle and propagation time difference values. Based on the camera signal arrival angle and propagation time difference values, three-dimensional spatial geometric positioning is performed. The position of the signal source in the three-dimensional coordinate system is solved by constructing a spatial geometric model. This process uses the center of the multi-antenna array as the origin and establishes an isosurface intersection model based on the direction vector described by the arrival angle and the time difference mapping distance characterized by the signal propagation time difference. The system converts the time difference of each pair of antennas into a signal propagation path length difference and plots it as a hyperboloid geometric equation. Simultaneously, the direction information provided by the arrival angle is represented as a ray direction constraint, and the intersection of these two is the estimated position of the target signal source. To ensure calculation accuracy, the system solves for the intersection set of each antenna pair and extracts the optimal coordinate result using the least squares error fitting method. Furthermore, it performs robust modeling for environmental occlusion and multipath reflection, eliminating outliers and anomalous jumps. Based on the obtained three-dimensional coordinate position, combined with the detector's current position and attitude angle parameters, the system calculates the spatial distance from the signal source to the detector, i.e., the signal source distance value, and links this distance value with data such as signal power.

[0027] In one specific embodiment, the process of performing step S2 may specifically include the following steps: The signal propagation path loss value is obtained by performing a logarithmic operation on the free space path loss based on the signal source distance value. Based on the current environmental loss factor values, the environmental-related attenuation loss values ​​are obtained through weighted calculation using environmental temperature correction coefficient, relative humidity correction coefficient, and spatial obstacle correction coefficient. The total channel attenuation loss is obtained by summing the signal propagation path loss and the environmental attenuation loss. The total channel attenuation loss value is added to the wireless signal received power value using power compensation calculation to obtain the wireless signal transmitted power value.

[0028] Specifically, the free space path loss formula is called using the signal source distance as an input parameter, and the propagation loss of the path is modeled using a logarithmic operation method. This process uses 20×log 10Expanded in the form (d), where d is the signal source distance in meters, this model reflects the power-law attenuation of signal energy in free space due to increased propagation distance. To accommodate the varying sensitivity of different frequencies to propagation distance, a frequency factor is introduced, combined with the specific center frequency of the operating band, to improve the model's fit to the propagation behavior of 2.4GHz and 5GHz signals in space. The basic path loss calculation yields the signal propagation path loss value without considering environmental interference, reflecting the energy loss range under pure geometric extension. In actual signal propagation, environmental factors such as temperature, humidity, and spatial obstacles affect the propagation characteristics of wireless signals. Therefore, the system introduces an environmental correlation correction mechanism to correct and supplement the aforementioned free space path loss value. Based on real-time acquired environmental parameter data, including ambient temperature measured by a temperature sensor, relative humidity obtained by a humidity sensor, and spatial obstacle density calculated by the obstacle detection module, the system configures corresponding correction factors for these three types of parameters using a built-in correction coefficient model: temperature correction coefficient, humidity correction coefficient, and obstacle correction coefficient. The system performs a product operation on each correction coefficient and the difference between its corresponding environmental parameter and the reference value, and then linearly weights the correction results using appropriate weighting factors to form an environmentally correlated attenuation loss value. This value quantifies the additional loss caused by non-ideal propagation factors in the real environment. The system then sums the two loss values—the signal propagation path loss value calculated based on the free-space logarithmic model and the environmentally correlated attenuation loss value calculated based on environmental parameters—to obtain the total channel attenuation loss value. This total loss value comprehensively reflects all energy attenuation encountered by the wireless signal from the transmitter to the receiver in the current space, including natural diffusion loss caused by geometric distance and non-ideal losses caused by environmental conditions. Based on this, the system performs a power compensation superposition operation on the total channel attenuation loss value and the received power value of the wireless signal received by the detector. The loss value is added as a compensation term to the received power to construct an estimated expression for the transmit power, and finally, the original transmit power value of the wireless signal is output.

[0029] In one specific embodiment, the process of performing step S3 may specifically include the following steps: The signal envelope of the wireless signal in the environment is extracted to obtain the wireless signal envelope waveform data; Based on the wireless signal envelope waveform data, the time distribution of active and silent segments of the signal is identified to obtain the time series data of the active segment. The time-domain parameters of the camera burst transmission mode are statistically analyzed on the time series data of the signal activity segment. The duration of the continuous signal activity segment is calculated as the transmission period parameter, the interval time between adjacent activity segments is calculated as the silence interval parameter, and the data packet length parameter is calculated by modulation characteristics and bit rate analysis to obtain the combination of camera time-domain characteristic parameters. A set of camera transmission features is constructed based on the combination of camera temporal feature parameters and power variance parameters.

[0030] Specifically, the system synchronously receives signals from the 2.4GHz and 5GHz bands via a radio frequency front-end module, and digitizes these signals using a high-resolution analog-to-digital converter, forming a continuous set of time-domain sampled data. Envelope extraction is then performed on this sampled data using Hilbert transform. An analytical signal is constructed from the original real-valued signal (the real part being the original signal itself, and the imaginary part being the result of its Hilbert transform). The modulus of this complex signal is then calculated to obtain wireless signal envelope waveform data reflecting the trend of signal energy intensity over time. Based on this wireless signal envelope waveform data, active and silent segments of the signal are identified by setting a dynamic detection threshold. This threshold is dynamically adjusted based on the mean and standard deviation of the envelope data within the current time window, specifically the range of the mean plus a certain multiple of the standard deviation, to accommodate the energy characteristics of active segments under different background noise conditions. The system defines regions in the envelope waveform that are continuously above this threshold as active segments and regions below the threshold as silent segments. The start time, end time, and duration of each active segment are recorded using timestamps as indices, along with the time interval between adjacent active segments, forming a time series of active signal segments. The system performs statistical analysis on the aforementioned time-series data to identify and quantify whether the wireless signal exhibits a typical camera transmission behavior pattern in the time domain. This behavior pattern is mainly characterized by several burst activity segments with relatively even durations and relatively fixed intervals, separated by significant silence periods. Its structural characteristics are clearly distinct from the continuous broadcasting of WiFi routers or the frequency-hopping short burst patterns of Bluetooth devices. Specifically, the system statistically averages the duration of each continuous signal activity segment as a transmission period parameter and averages the silence interval between adjacent activity segments to obtain the silence interval parameter. Simultaneously, to calculate the data packet length parameter, the system further analyzes the modulation characteristics and bit rate within each activity segment. This process combines signal modulation identification algorithms, such as higher-order cumulant analysis and spectral envelope morphology matching, to determine the current signal modulation scheme, such as QPSK, 16QAM, or higher-order modulation forms. Then, it combines the sampling rate and symbol rate to deduce the bit transmission capacity per unit time, thereby calculating the data packet length contained in each activity segment as the product of the transmission duration and the bit rate. The system combines three time-domain parameters—transmission period, silence interval, and data packet length—to form a set of camera time-domain feature parameters that represent the current signal burst structure. Based on this set of camera time-domain feature parameters, discriminative information from the power statistics dimension is introduced. Specifically, within each signal activity segment, variance calculation is performed on the received power sequence of the sampled signal to obtain a power variance parameter. This parameter reflects whether the output power at the signal transmitter is stable. This power statistics parameter is then combined with the three aforementioned time-domain structure parameters to form a set of camera transmission features. This feature set is represented as a four-dimensional vector, with each dimension corresponding to the transmission period, silence interval, data packet length, and power variance, respectively.

[0031] In one specific embodiment, the process of performing the step of identifying the time distribution of active and silent segments of a signal based on the wireless signal envelope waveform data to obtain the time series data of the active segment can specifically include the following steps: Calculate the mean and standard deviation of the wireless signal envelope waveform data respectively; The detection threshold for the active segment of the signal is set by adding the product of the mean value, the adjustment coefficient, and the standard deviation value. Based on the signal activity segment detection threshold, the amplitude determination calculation is performed on the wireless signal envelope waveform data point by point. When the envelope amplitude exceeds the detection threshold, it is identified as a signal activity state, and when it is below the detection threshold, it is identified as a signal silence state, thus obtaining the signal activity segment time series data.

[0032] Specifically, the mean and standard deviation of the wireless signal envelope waveform data are calculated separately. The mean represents the overall amplitude level of the signal within the observation time window and is a key parameter for determining the baseline of signal energy distribution. The standard deviation measures the dispersion of each sampling point relative to the mean, thus reflecting the local volatility or activity level of the signal. Together, they constitute the statistical feature space of the envelope data. To identify active and silent segments in the envelope waveform, an amplitude threshold for classification is set. This threshold is dynamically generated based on the current data distribution characteristics. The system sets an adjustment coefficient as a hyperparameter, a floating-point value between 1 and 4, which is optimized based on experience or training data. This coefficient is multiplied by the standard deviation and then superimposed on the mean to generate the signal activity segment detection threshold for the current envelope waveform. Once the detection threshold is set, the system performs amplitude judgment calculations on the envelope waveform data point by point. For each sampling time t, the system compares the envelope amplitude value corresponding to that point with a set threshold. If the amplitude value is greater than the detection threshold, the system considers the current time to be in a signal active state; if the amplitude value is less than or equal to the detection threshold, the system considers the current time to be in a signal silent state. Through point-by-point discrimination, the system divides the envelope data on the entire time axis into alternating active and silent intervals. After completing the state identification of all sampling points, the system merges the data points consecutively identified as "active" in chronological order, marking their start time, end time, and duration. It also inserts timestamps marking adjacent silent intervals between different active segments, thus forming a time series data of signal active segments.

[0033] In one specific embodiment, the process of performing step S4 may specifically include the following steps: The probability density function of the camera signal state and the probability of the first camera signal state are calculated by using a binary state transition probability model on the camera transmission feature set to obtain the probability of the first camera signal state and the probability of the first non-camera signal state. Based on the state probabilities of the first camera signal and the first non-camera signal, and combined with the state probabilities of the current time and the Markov chain transformation matrix, a Bayesian posterior probability update operation is performed to obtain the state probabilities of the second camera signal and the second non-camera signal. Signal state transition probability values ​​are generated based on the probability of the second camera signal state and the probability of the second non-camera signal state.

[0034] Specifically, the camera transmission feature set includes multiple dimensions such as signal transmission period, silence interval, data packet length, and power variance, which constitute a vector space with temporal behavioral characteristics. The system inputs this vector into a binary state transition probability model. The model uses "camera signal state" and "non-camera signal state" as two mutually exclusive output classification labels and estimates the probability density of the input features' distribution in these two states. The system models camera signal samples and non-camera samples separately based on a Gaussian mixture model. During training, the EM algorithm (Expectation Maximization) is used to fit the feature mean, covariance matrix, and weight parameters for each class. In the inference phase, the system uses the input feature vector as the observation variable and calculates the conditional probability density function value of the two models at the current feature point, outputting the probability of the first camera signal state and the probability of the first non-camera signal state. These two probability values ​​are combined to form the prior state judgment result at the current moment. The probability of the first camera signal state represents the degree of fit of the current feature to the camera model, while the probability of the first non-camera signal state reflects its degree of fit to the background signal or legitimate communication device model. The system introduces a Markov chain structure, combining the prior judgment of the current state with historical state transition patterns to establish a Bayesian posterior probability update mechanism. During the initialization phase, the system statistically analyzes the state transition frequency between camera signal states and non-camera signal states based on historical signal data, constructing a state transition probability matrix. This matrix contains the conditional probabilities of transitioning from the previous state to the current state, such as the probability of transitioning from "camera state" to "non-camera state," and the probability of maintaining the "non-camera state" back to "non-camera state." In actual computation, the system calls the current state transition matrix and combines it with historical state distribution information to perform Bayesian posterior correction on the current first state probability value. Specifically, it uses the first camera state probability output by the model as the likelihood function, the previous state probability as the prior distribution, and performs weighted fusion using Markov state transition weights to finally calculate the second camera signal state probability and the second non-camera signal state probability. After the completion of the verification update, the system generates a signal state transition probability value based on the state probability results of the second stage. This probability value is interpreted as an expression of the confidence that the signal will transition from a "non-camera state" to a "camera state" at the current moment, represented as a continuous quantity between 0 and 1. To generate this probability, the system introduces a time decay factor, which gives higher weight to the probability of historical states that are close in time to the current decision, while the influence of states at more distant time points gradually decays, thus forming a state estimation sequence that is adaptively updated within a time window. This signal state transition probability value is combined with an information entropy function for confidence assessment. By calculating the entropy value of the state probability distribution, the system reflects the degree of certainty in the decision-making process. The lower the entropy value, the more concentrated and stable the current judgment is; conversely, the higher the entropy value, the greater the uncertainty in the judgment.

[0035] In one specific embodiment, the process of performing the step of calculating the probability density function of the camera signal state and the non-camera signal state on the camera transmission feature set using a binary state transition probability model to obtain the first camera signal state probability and the first non-camera signal state probability can specifically include the following steps: The camera transmission feature set is input into the Gaussian mixture model corresponding to the camera signal state and the non-camera signal state in the binary state transition probability model to calculate the probability density function, and the result of the two-state Gaussian mixture probability density calculation is obtained. Based on the calculation results of the bi-state Gaussian mixture probability density, the probability density values ​​under the camera signal state and the probability density values ​​under the non-camera signal state are calculated respectively to obtain the probability of the first camera signal state and the probability of the first non-camera signal state.

[0036] Specifically, the camera transmission feature set includes parameters such as transmission period, silence interval, data packet length, and power variance. Each parameter represents the behavioral characteristics of the wireless signal in the time domain, amplitude domain, or modulation structure. These parameters, when combined, form a multi-dimensional feature vector used as input to the classification model. During model training, the system uses a large amount of labeled sample data to model the feature vectors for both "camera signal state" and "non-camera signal state" categories. Each model employs a Gaussian Mixture Model (GMM) for distribution fitting. GMM can express complex non-unimodal feature spaces through weighted combinations of multiple Gaussian distribution components, making it suitable for handling diverse feature variations in camera signals. The system optimizes the parameters in the GMM model using the Expectation-Maximization (EM) algorithm, specifically including the mean vector, covariance matrix, and weighting factors for each Gaussian component. These parameters characterize the concentration, diffusion, and distribution ratio of features in different distribution regions. During the recognition phase, after acquiring a new camera transmission feature set, the system inputs this feature set into the two pre-trained Gaussian Mixture Models to calculate the probability density function. The system substitutes the input feature vector into the Gaussian Mixture Model (GMM) under camera signal conditions, sequentially traversing all K Gaussian distribution components in the model. For each component, the system calculates the probability density value of the input vector under that Gaussian distribution and multiplies it by the component's weighting factor. The weighted density values ​​of all components are then summed to obtain the comprehensive probability density value of the feature under camera signal conditions. The system similarly inputs the feature vector into the GMM under non-camera signal conditions, calculating and weighting the probability density value of the feature under non-camera signal conditions using the Gaussian density functions of each component. This process completes the numerical solution based on the dual-state Gaussian mixture probability density function, obtaining probability density calculation results for two states, corresponding to the degree of fit of the target signal in the states of "belonging to camera signal" and "belonging to non-camera signal," respectively. These two probability density results are normalized so that their sum equals 1 to meet the probability consistency requirement of the state space. The system uses the probability density of the camera state as the numerator and the sum of the two density values ​​(camera state and non-camera state) as the denominator to construct a Bayesian conditional probability expression, thus obtaining the first camera signal state probability, which is the normalized probability that the input feature belongs to the camera signal. Simultaneously, its complementary value is the first non-camera signal state probability. If the density of the camera state is much greater than that of the non-camera state, the first camera state probability is close to 1, indicating high reliability. Conversely, if the densities of the two states are similar, the state probability is close to neutral, indicating that the current feature sample belongs to the boundary sample, and the reliability of the judgment result is low.

[0037] In one specific embodiment, the process of performing a Bayesian posterior probability update operation based on the state probabilities of the first camera signal and the first non-camera signal, combined with the current state probability and the Markov chain transformation matrix, to obtain the state probabilities of the second camera signal and the second non-camera signal, can specifically include the following steps: Obtain the prior probability of the camera signal state and the prior probability of the non-camera signal state at the current moment from the historical state records; Based on the prior probabilities of camera signal state and non-camera signal state, the transition probabilities from camera signal state to camera signal state, from camera signal state to non-camera signal state, from non-camera signal state to camera signal state, and from non-camera signal state to non-camera signal state are extracted from the Markov chain transition matrix to obtain a quaternary state transition probability matrix. The probability of the first camera signal state and the probability of the first non-camera signal state are used as likelihood probabilities. Bayesian operations are then performed on the quaternary state transition probability matrix to obtain the probability of the second camera signal state and the probability of the second non-camera signal state.

[0038] Specifically, the prior probabilities of the camera signal state and non-camera signal state at the current moment are obtained from historical state records. The system then enters the state transition probability acquisition phase. This process is based on a pre-trained and stored Markov chain state transition matrix, which describes the transition relationship between signal states at two adjacent moments, including the probability of transitioning from "camera signal state" to "camera signal state" (denoted as P). 11 The probability (P) of transitioning from "camera signal state" to "non-camera signal state". 12 The probability (P) of transitioning from a "non-camera signal state" to a "camera signal state". 21 ), and the probability of remaining in the "non-camera signal state" from the "non-camera signal state" (P 22These four probabilities constitute a quaternary state transition probability matrix, defining the transition possibilities between any combination of states. Each row corresponds to a current state, and each column corresponds to a next state. The sum of the probabilities in each row of the matrix should be 1, satisfying the consistency of the total probability. The system can obtain this quaternary transition matrix based on the current prior state by looking up a table or directly calling the parameter model. In obtaining the state prior probabilities and the quaternary transition matrix, the system integrates the probability of the first camera signal state and the probability of the first non-camera signal state at the current moment as the likelihood probabilities from the observation model. In specific operations, the Bayesian operation mechanism performs a weighted product of the historical state prior and the current observation likelihood, and obtains the updated state posterior probability through a normalization operation. This process is expressed as: the posterior probability of "camera signal state" at the current moment is equal to the prior probability of "camera signal state" at the previous moment multiplied by P. 11 Then multiply by the probability density value of the current observed feature falling into the "camera signal model"; add the prior probability of being in the "non-camera signal state" at the previous time step multiplied by P. 21 Then, multiply by the probability density value of the current observation under the "camera signal model", and finally divide by the standardized denominator, which is the sum of the overall joint probability densities, to obtain the probability of the second camera signal state. Similarly, the probability of the second non-camera signal state is obtained symmetrically through the joint calculation of the prior, transition, and observation components.

[0039] In one specific embodiment, the process of performing step S5 may specifically include the following steps: The signal state transition probability value is multiplied by the first weighting coefficient, and then the product of the wireless signal transmission power value and the second weighting coefficient is added to obtain the comprehensive fusion decision score. The integrated decision score is compared with the camera detection decision threshold to output the probability decision value of the existence of the hidden camera signal; The probability judgment value of the presence of hidden camera signals is continuously verified within a time window. The frequency of the camera signal presence judgment within the time window is counted. When the frequency exceeds the preset consistency verification threshold, the hidden camera detection result is confirmed.

[0040] Specifically, the signal state transition probability value represents the credibility of the current feature switching between "camera signal state" and "non-camera signal state." It is derived from previous derivations using Bayesian inference and Markov state chain models, exhibiting temporal evolution characteristics. The wireless signal transmission power value, on the other hand, is derived from the inverse calculation of the received signal power in the environment after distance compensation and channel loss correction. It reflects the physical transmission intensity level of the signal source, serving as a direct physical clue to determine whether it belongs to a low-power covert device (such as a pinhole camera). The system sets two independent weighting coefficients, corresponding to the weight contribution values ​​of the state discrimination dimension and the energy dimension, respectively. The first weighting coefficient is used to adjust the influence ratio of the state probability, while the second weighting coefficient is used to control the pulling effect of the transmission power value on the result. In the fusion calculation stage, the system multiplies the signal state transition probability value with the first weighting coefficient and simultaneously multiplies the wireless signal transmission power value with the second weighting coefficient. These two products are then added together to obtain the fused comprehensive decision score. This score represents a fusion score indicating whether the current signal is emitted by a hidden camera. A higher score indicates that the signal closely resembles the standard behavior model of a hidden device in terms of both state and power characteristics. To convert this fusion score into a discrete logic judgment result, the system introduces a camera detection decision threshold as a recognition boundary in the score domain. This threshold is dynamically configured based on factors such as environmental noise level, false alarm tolerance, and detection sensitivity, or automatically adjusted during model training based on ROC curve analysis. During operation, the system compares the current fusion score with the camera detection decision threshold. If the fusion score is greater than or equal to the threshold, a probability judgment value indicating the presence of a hidden camera signal is output, with a value of 1 or close to 1. If the fusion score is lower than the threshold, the output value is 0 or close to 0. Due to short-term interference, momentary misjudgments, or overlapping signal features in the wireless signal environment, the system introduces a time window mechanism to statistically analyze the decision values ​​at multiple consecutive time points to verify the temporal stability and consistency of the signal. This mechanism operates in the form of a sliding window, setting a fixed-length time window, such as 5 seconds or 50 sampling periods, and recording the probability judgment value generated in each decision period within this window. The system accumulates and statistically analyzes these judgment values, calculating the frequency of the "hidden camera signal exists" judgment within this time period, i.e., the number of times the judgment value is 1 or exceeds a certain confidence threshold, and compares this frequency with a system-preset consistency verification threshold. When the frequency is greater than or equal to the consistency threshold, for example, 35 out of 50 judgments indicate presence, the system confirms that the current signal source has a high probability of being a hidden camera, and finally outputs the hidden camera detection result, marking this conclusion as a credible judgment state.During this confirmation process, the system incorporates the continuity characteristics of the frequency distribution, such as whether there are three or more consecutive positive judgment results or whether the judgment signal appears stably in multiple sub-intervals, to enhance its ability to distinguish critical situations and improve the system's decision sensitivity in edge states. To adapt to changes in signal interference intensity in different scenarios, the system dynamically adjusts the time window length and consistency threshold to achieve an adaptive balance between sensitivity and stability.

[0041] The detection method based on the detector's wireless signal in the embodiments of the present invention has been described above. The detector device in the embodiments of the present invention is described below. Please refer to [link / reference]. Figure 2 One embodiment of the detector device in this invention includes: The acquisition module 001 is used to acquire the wireless signal received power value and the signal source distance value through the detector; The compensation module 002 is used to perform channel attenuation compensation on the wireless signal received power value based on the signal source distance value to obtain the wireless signal transmitted power value; The identification module 003 is used to perform signal envelope detection and transmission feature identification on wireless signals in the environment to obtain a set of camera transmission features; The calculation module 004 is used to perform probability calculations of camera signal state and non-camera signal state based on the camera transmission feature set, and to obtain the signal state transition probability value. The generation module 005 is used to perform weighted fusion decision based on the signal state transition probability value and the wireless signal transmission power value to generate the hidden camera detection result.

[0042] Through the collaborative efforts of the aforementioned components, and by establishing a channel attenuation compensation model incorporating free-space path loss and environmental loss factors, the true transmit power of the wireless signal can be accurately reconstructed. This eliminates the impact of distance and environmental factors on detection accuracy, significantly improving the accuracy of power decision-making compared to traditional fixed-threshold detection methods. Employing synchronous reception technology in the 2.4GHz and 5GHz bands, combined with a software-defined radio architecture, comprehensive coverage of the camera device's operating frequency band is achieved, avoiding the limitations of single-band detection and enhancing the system's spectrum coverage and signal acquisition efficiency. Based on real-time monitoring of environmental parameter sensor arrays and dynamic calculation of environmental loss factors, the detection system achieves adaptive adjustment to different environmental conditions, overcoming the performance degradation problem of traditional methods in complex environments and enhancing the system's environmental adaptability and robustness. Through signal phase difference analysis and signal propagation time difference analysis of the three-antenna array, precise three-dimensional positioning of the signal source is achieved, providing accurate spatial information for subsequent distance estimation and power compensation, thus improving the overall positioning accuracy and reliability of the detection system. By analyzing the unique burst transmission patterns of camera signals and extracting temporal feature parameters such as transmission period, silence interval, and data packet length, the system can effectively distinguish camera signals from signals from other wireless devices such as WiFi and Bluetooth, significantly reducing false alarm and false negative rates. Employing a Markov chain framework and Bayesian inference algorithm, an intelligent probabilistic inference mechanism is established, which can fully utilize historical state information and current observation features for dynamic probability updates, exhibiting stronger anti-interference capabilities and decision accuracy compared to simple threshold decision methods. By weighted fusion of state transition probability values ​​and power compensation information, multi-dimensional information is comprehensively utilized, overcoming the limitations of single-feature decision-making and improving the overall decision-making capability and reliability of the detection system. Continuous statistical verification within a sliding time window ensures the temporal consistency and stability of detection results, effectively suppressing the impact of transient interference and noise on the detection results, and improving the system's anti-interference capability and detection reliability.

[0043] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0044] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0045] The above-described embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A detection method based on the wireless signal of a detector, characterized in that, include: The detector acquires the wireless signal reception power value and the signal source distance value; Based on the signal source distance value, channel attenuation compensation is performed on the wireless signal received power value to obtain the wireless signal transmitted power value; Signal envelope detection and transmission feature identification are performed on wireless signals in the environment to obtain a set of camera transmission features; Based on the set of camera transmission features, perform probability calculations on camera signal states and non-camera signal states to obtain signal state transition probability values; The hidden camera detection result is generated by weighted fusion decision based on the signal state transition probability value and the wireless signal transmission power value.

2. The detection method based on detector wireless signals according to claim 1, characterized in that, The process of acquiring the wireless signal received power value and signal source distance value through the detector includes: The system synchronously receives wireless signals in the 2.4GHz and 5GHz bands in the environment to obtain multi-band camera wireless signal data and corresponding wireless signal received power values, while simultaneously measuring the received signal strength indicator values ​​for each band. The detector acquires ambient temperature, relative humidity, and spatial obstacle density values ​​based on its built-in environmental parameter sensor array, and calculates the current environmental loss factor value. The signal phase difference analysis and signal propagation time difference analysis are performed on the wireless signal data of the multi-band camera, and the propagation delay is corrected by combining the current environmental loss factor value to obtain the signal arrival angle value and the signal propagation time difference value. Three-dimensional spatial geometric positioning is performed based on the camera signal arrival angle value and the signal propagation time difference value to obtain the signal source distance value in the environment.

3. The detection method based on the detector's wireless signal according to claim 2, characterized in that, The step of performing channel attenuation compensation on the wireless signal received power value based on the signal source distance value to obtain the wireless signal transmitted power value includes: The signal propagation path loss value is obtained by performing a logarithmic operation on the free space path loss based on the signal source distance value. Based on the current environmental loss factor value, the environmental-related attenuation loss value is obtained by weighted calculation using the environmental temperature correction coefficient, relative humidity correction coefficient, and spatial obstacle correction coefficient. The total channel attenuation loss is obtained by summing the signal propagation path loss value and the environment-related attenuation loss value. The total channel attenuation loss value is added to the wireless signal received power value to obtain the wireless signal transmitted power value.

4. The detection method based on detector wireless signals according to claim 1, characterized in that, The process of performing signal envelope detection and transmission feature recognition on wireless signals in the environment to obtain a set of camera transmission features includes: The signal envelope of the wireless signal in the environment is extracted to obtain the wireless signal envelope waveform data; Based on the wireless signal envelope waveform data, the time distribution of the active and silent segments of the signal is identified to obtain the time series data of the active segments. The time-domain parameters of the camera burst transmission mode are statistically analyzed on the time series data of the signal activity segment. The duration of the continuous signal activity segment is calculated as the transmission period parameter, the interval time between adjacent activity segments is calculated as the silence interval parameter, and the data packet length parameter is calculated by modulation characteristics and bit rate analysis to obtain the combination of camera time-domain characteristic parameters. A set of camera transmission features is constructed based on the combination of camera temporal feature parameters and power variance parameters.

5. The detection method based on detector wireless signals according to claim 4, characterized in that, The step of identifying the time distribution of active and silent segments of the signal based on the wireless signal envelope waveform data to obtain time series data of the active segments includes: Calculate the mean and standard deviation of the wireless signal envelope waveform data respectively; The signal activity segment detection threshold is set by adding the mean value to the product of the adjustment coefficient and the standard deviation value; Based on the signal activity segment detection threshold, the amplitude determination calculation is performed on the wireless signal envelope waveform data point by point. When the envelope amplitude exceeds the detection threshold, it is identified as a signal activity state, and when it is below the detection threshold, it is identified as a signal silence state, thus obtaining the signal activity segment time series data.

6. The detection method based on detector wireless signals according to claim 1, characterized in that, The step of performing probability calculations on camera signal states and non-camera signal states based on the camera transmission feature set to obtain signal state transition probability values ​​includes: The probability density function of the camera signal state and the probability of the first camera signal state are calculated by using a binary state transition probability model on the camera transmission feature set to obtain the probability of the first camera signal state and the probability of the first non-camera signal state. Based on the state probability of the first camera signal and the state probability of the first non-camera signal, a Bayesian posterior probability update operation is performed by combining the state probability at the current moment and the Markov chain transformation matrix to obtain the state probability of the second camera signal and the state probability of the second non-camera signal. Signal state transition probability values ​​are generated based on the second camera signal state probability and the second non-camera signal state probability.

7. The detection method based on detector wireless signals according to claim 6, characterized in that, The step of calculating the probability density function of the camera signal state and the non-camera signal state on the camera transmission feature set using a binary state transition probability model to obtain the first camera signal state probability and the first non-camera signal state probability includes: The camera transmission feature set is input into the Gaussian mixture model corresponding to the camera signal state and the non-camera signal state in the binary state transition probability model to calculate the probability density function, and the result of the dual-state Gaussian mixture probability density calculation is obtained. Based on the calculated results of the dual-state Gaussian mixture probability density, the probability density values ​​under the camera signal state and the probability density values ​​under the non-camera signal state are calculated respectively to obtain the probability of the first camera signal state and the probability of the first non-camera signal state.

8. The detection method based on detector wireless signals according to claim 7, characterized in that, The step of performing a Bayesian posterior probability update operation based on the state probabilities of the first camera signal and the first non-camera signal, combined with the current state probability and the Markov chain transformation matrix, to obtain the state probabilities of the second camera signal and the second non-camera signal includes: Obtain the prior probability of the camera signal state and the prior probability of the non-camera signal state at the current moment from the historical state records; Based on the prior probabilities of the camera signal state and the prior probabilities of the non-camera signal state, the transition probabilities from camera signal state to camera signal state, from camera signal state to non-camera signal state, from non-camera signal state to camera signal state, and from non-camera signal state to non-camera signal state are extracted from the Markov chain transition matrix to obtain a quaternary state transition probability matrix. The probability of the first camera signal state and the probability of the first non-camera signal state are used as likelihood probabilities. Bayesian operations are then performed on the quaternary state transition probability matrix to obtain the probability of the second camera signal state and the probability of the second non-camera signal state.

9. The detection method based on detector wireless signals according to claim 1, characterized in that, The step of generating a hidden camera detection result by weighted fusion decision based on the signal state transition probability value and the wireless signal transmission power value includes: The product of the signal state transition probability value and the first weighting coefficient is added to the product of the wireless signal transmission power value and the second weighting coefficient to obtain the comprehensive fusion decision score. The integrated decision score is compared with the camera detection decision threshold, and the probability decision value of the existence of the hidden camera signal is output. The probability judgment value of the presence of the hidden camera signal is continuously verified within a time window. The frequency of the camera signal presence judgment within the time window is counted. When the frequency exceeds the preset consistency verification threshold, the hidden camera detection result is confirmed.

10. A detector device, characterized in that, For performing the detection method based on detector wireless signals as described in any one of claims 1-9, the detector device comprises: The acquisition module is used to acquire the wireless signal received power value and the signal source distance value through the detector; The compensation module is used to perform channel attenuation compensation on the wireless signal received power value based on the signal source distance value to obtain the wireless signal transmitted power value; The identification module is used to detect the signal envelope and identify the transmission features of wireless signals in the environment to obtain a set of camera transmission features. The calculation module is used to perform probability calculations of camera signal state and non-camera signal state based on the camera transmission feature set, and obtain signal state transition probability values; The generation module is used to perform weighted fusion decision based on the signal state transition probability value and the wireless signal transmission power value to generate the hidden camera detection result.

Images